PRISMA flow diagram errors are among the most common reasons for revision requests in systematic review manuscripts. Peer reviewers and journal editors frequently identify inconsistencies, missing information, and incorrect template usage that undermine the transparency of the study selection process. Understanding these common mistakes before creating your PRISMA diagram can save significant revision time and improve the quality of your reporting.
This guide identifies the most frequent PRISMA 2020 flow diagram errors encountered in peer review and provides practical solutions for each.
Mistake 1: Using the Wrong Template Type
The problem: Many researchers use a databases-only template when they also searched citation lists, websites, or grey literature, or vice versa. Others use a "new review" template for an updated systematic review.
The fix: Choose your template based on two questions: (1) Is this a new or updated review? (2) Did you search only databases/registers, or also other sources? If you used citation searching, that counts as "other sources." Read our detailed guide on choosing the right PRISMA template.
Mistake 2: Numbers Don't Add Up
The problem: The most flagged error in peer review. For example, records screened minus records excluded should equal reports sought for retrieval. If your numbers are inconsistent, reviewers will question the accuracy of your entire selection process.
The fix: Work through the diagram from top to bottom and verify:
- Records identified minus duplicates/removals = records screened
- Records screened minus excluded = reports sought for retrieval
- Reports sought minus not retrieved = reports assessed for eligibility
- Reports assessed minus excluded = included reports
Our free PRISMA flow diagram generator displays numbers in real time so you can verify consistency as you enter data.
Mistake 3: Missing or Vague Exclusion Reasons
The problem: Stating "did not meet inclusion criteria" as the only exclusion reason at the eligibility stage, or providing only one or two generic reasons when dozens of studies were excluded.
The fix: Provide specific, mutually exclusive exclusion reasons with counts for each. Use categories like "Wrong population," "Wrong intervention," "Wrong study design," "Insufficient data," or "Duplicate cohort." Each excluded report should be assigned to exactly one primary reason.
Mistake 4: Confusing Records, Reports, and Studies
The problem: Using "studies" throughout the diagram when PRISMA 2020 distinguishes between records (database entries), reports (full-text documents), and studies (research projects). One study can produce multiple reports.
The fix: Use "records" for the identification and screening phases, "reports" for the retrieval and eligibility phases, and "studies" only in the final included box. PRISMA 2020 explicitly defines these terms, so follow the terminology in the PRISMA 2020 checklist.
Mistake 5: Still Using the PRISMA 2009 Format
The problem: Submitting a flow diagram based on the original 2009 PRISMA template, which lacks the pre-screening removal box, doesn't distinguish between databases and other sources, and uses outdated terminology.
The fix: Update to the PRISMA 2020 format. Most journals now expect PRISMA 2020 compliance. Use our online PRISMA flow diagram tool to create a diagram that follows the current standard. For a detailed comparison, see PRISMA 2020 vs original PRISMA.
Mistake 6: Omitting the Pre-Screening Removal Box
The problem: Not reporting how many duplicate records, automation-removed records, and other pre-screening removals occurred. PRISMA 2020 added this box specifically to capture these numbers.
The fix: Always include the "Records removed before screening" box, even if the only removal was deduplication. Report duplicates, automation tool removals, and other reasons separately.
Mistake 7: Not Reporting Database-Specific Counts
The problem: Reporting only a total number of records without breaking down results by database.
The fix: Report the number of records returned by each database separately in the identification box (e.g., "PubMed: 1,203; Scopus: 892; Web of Science: 752"). This allows readers to assess the contribution of each source.
Mistake 8: Inconsistent Numbers Between Text and Diagram
The problem: The numbers in the PRISMA flow diagram don't match the numbers reported in the results section of the manuscript text.
The fix: Cross-reference every number in your flow diagram with the corresponding number in your text after completing the manuscript. This should be part of your final pre-submission check.
Mistake 9: Missing the "Other Sources" Column
The problem: Reporting citation searching or grey literature sources in the text but not including them in the flow diagram because a databases-only template was used.
The fix: If you searched any source beyond databases and registers, use a template that includes the "other sources" column. Citation searching, contacting organizations, and searching websites all count as other sources.
Mistake 10: Not Using a Standardized Tool
The problem: Drawing the PRISMA diagram manually in PowerPoint, Word, or a general diagramming tool, resulting in non-standard layouts, missing boxes, or incorrect flow.
The fix: Use a dedicated PRISMA 2020 flow diagram generator that enforces the correct structure. Our PRISMA 2020 diagram maker automatically creates properly formatted diagrams from your data input.
Prevention Checklist
Before submitting your manuscript, verify:
- Correct template type selected (new/updated, databases only/other sources)
- All numbers flow logically from identification to inclusion
- Specific exclusion reasons provided with counts at eligibility stage
- Correct use of "records," "reports," and "studies" terminology
- Pre-screening removal box included with breakdown
- Database-specific counts in identification box
- Numbers match between flow diagram and manuscript text
- PRISMA 2020 format used (not PRISMA 2009)
Frequently Asked Questions
What is the most common PRISMA diagram mistake?
The most frequently identified error in peer review is numerical inconsistency, meaning the numbers in different boxes of the flow diagram don't add up logically. For example, if 1,000 records were screened and 800 were excluded, then 200 should appear in the next box, not some other number.
Will journals reject my paper for PRISMA errors?
Most journals will request revisions rather than immediate rejection for PRISMA errors. However, significant errors (using the wrong PRISMA version, major numerical inconsistencies, missing flow diagram entirely) can lead to desk rejection at some journals.
Can I use different exclusion reasons than the standard categories?
Yes. Exclusion reasons should be specific to your review topic. While common categories include wrong population, wrong intervention, wrong outcome, and wrong study design, you should use whatever categories accurately describe why studies were excluded from your specific review.
How do I handle studies with multiple exclusion reasons?
Assign each excluded report to one primary exclusion reason. If a report failed multiple criteria, use the most important or first-encountered reason. Be consistent in your approach and document your decision rule.